# The following function calculates the Agglomeration metric for the given graph

def agglomeration(g):
    agglomeration_numerator = 0.0
    total_degrees = 0.0
    
    degrees = nx.degree(g)
    for edge in g.edges():
        agglomeration_numerator += degrees[edge[0]] * degrees[edge[1]]
        
    for item in degrees:
        total_degrees += degrees[item]
        
    return agglomeration_numerator / pow(total_degrees, 2)

# The following function calculates the degree distribution histogram for the given graph

def degree_distribution(g):
    hist = {}
    for n in g.nodes():
        degree = g.degree(n)
        if degree not in hist:
            hist[degree] = degree
        else:
            hist[degree] = hist[degree]+degree
            
    return hist

# The following code loads up the graphml files calculates the necessary metrics
# and generates visualizations and degree distribution histograms.

import networkx as nx
from networkx.drawing.nx_agraph import graphviz_layout
from multiprocessing import Process
from multiprocessing import Pool
import threading
import sys
import os
import matplotlib.pyplot as plt
import numpy as np

%matplotlib inline  

final_mids = []
mids = os.listdir("graphml")
for i in mids:
    if(i.endswith("graphml")):
        final_mids.append(i)

for f in final_mids:
    print("\n\n\n")
    print("Reading graphml/" + f)
    g = nx.read_graphml("./graphml/" + f)
    
    dd = degree_distribution(g)
    
    print("Clustering Coefficient: " + str(nx.average_clustering(g.to_undirected(), weight='weight')))
    print("Avg Path Length: " + str(nx.average_shortest_path_length(g)))
    print("Agglomeration: " + str(agglomeration(g)))
    print("Number of nodes: "+ str(len(g.nodes())))
    print("Number of edges: "+ str(len(g.edges())))
    print("Unique degrees: " + str(len(dd)))
    
    bet_cen=nx.betweenness_centrality(g, weight='weight')
    bet_color=bet_cen.items()
    bet_color.sort()
    bet_color=[(b) for (a,b) in bet_color]
    
    eig_cen=nx.eigenvector_centrality(g, max_iter=500000, weight='weight')
    eig_size=eig_cen.items()
    eig_size.sort()
    eig_size=[((b)*2000)+20 for (a,b) in eig_size]
   
    spring_pos=nx.spring_layout(g,iterations=1000, weight='weight', dim=2)
    pos = graphviz_layout(g, root=g.node['null'])
    plt.figure(figsize=(10,10))
    nx.draw(g, pos=spring_pos, node_color=bet_color, node_size=eig_size, cmap=plt.cm.Greens, edge_weight="weight", with_labels=False)
    plt.show()
    
    D = dd
    print("Degree Distribution: " + str(dd))
    print("\n")

    plt.title("Degree histogram for " + f + "\n")
    plt.bar(range(len(D)), D.values(), align='center')
    tick_marks = np.arange(len(D.values()))
    plt.ylabel('Frequency')
    plt.xlabel('Degrees')
    plt.show()



Reading graphml/292jsb0552.midi.graphml
Clustering Coefficient: 0.0229917122188
Avg Path Length: 3.15738553417
Agglomeration: 0.158601800931
Number of nodes: 275
Number of edges: 2088
Unique degrees: 55

Degree Distribution: {2: 120, 3: 30, 4: 112, 5: 80, 6: 96, 7: 35, 8: 104, 9: 45, 10: 30, 11: 66, 12: 84, 13: 78, 14: 56, 15: 60, 16: 64, 17: 136, 18: 18, 19: 133, 20: 40, 21: 42, 22: 88, 23: 23, 24: 120, 25: 75, 26: 130, 27: 108, 28: 112, 29: 116, 31: 31, 32: 96, 33: 33, 35: 35, 36: 72, 37: 37, 38: 76, 39: 39, 40: 40, 42: 84, 43: 43, 45: 90, 46: 92, 47: 141, 50: 50, 54: 108, 60: 60, 61: 61, 63: 63, 66: 132, 67: 67, 74: 74, 75: 75, 77: 77, 80: 80, 96: 96, 123: 123}



Reading graphml/50_cent-in_da_club.midi.graphml
Clustering Coefficient: 0.0120648262587
Avg Path Length: 3.5752688172
Agglomeration: 0.207971807253
Number of nodes: 63
Number of edges: 282
Unique degrees: 21

Degree Distribution: {2: 26, 3: 18, 4: 28, 5: 10, 6: 24, 7: 21, 9: 9, 10: 50, 11: 55, 12: 24, 34: 34, 14: 14, 15: 30, 16: 32, 17: 17, 19: 38, 20: 20, 21: 42, 28: 28, 13: 13, 31: 31}



Reading graphml/acdc-dirty_deeds_done_dirt_cheap.mid.graphml
Clustering Coefficient: 0.00354762966973
Avg Path Length: 3.57131147541
Agglomeration: 0.227382147921
Number of nodes: 61
Number of edges: 184
Unique degrees: 17

Degree Distribution: {2: 56, 3: 18, 4: 20, 5: 10, 6: 18, 39: 39, 8: 24, 9: 9, 10: 10, 11: 11, 12: 24, 13: 39, 16: 16, 17: 17, 19: 19, 7: 7, 31: 31}



Reading graphml/ace_of_base-all_that_she_wants.mid.graphml
Clustering Coefficient: 0.0155058346985
Avg Path Length: 3.76989359241
Agglomeration: 0.134799628865
Number of nodes: 132
Number of edges: 438
Unique degrees: 20

Degree Distribution: {2: 74, 3: 48, 4: 84, 5: 55, 6: 48, 7: 14, 8: 72, 9: 27, 10: 70, 11: 22, 12: 24, 13: 13, 14: 42, 15: 30, 17: 17, 40: 40, 56: 56, 25: 25, 28: 84, 31: 31}



Reading graphml/adele-rolling_deep.midi.graphml
Clustering Coefficient: 0.0258326620925
Avg Path Length: 4.6290684624
Agglomeration: 0.0684676102133
Number of nodes: 243
Number of edges: 809
Unique degrees: 28

Degree Distribution: {2: 164, 3: 63, 4: 96, 5: 85, 6: 120, 7: 63, 8: 104, 9: 99, 10: 100, 11: 22, 12: 60, 13: 52, 14: 28, 15: 60, 16: 32, 17: 17, 18: 36, 19: 57, 20: 20, 21: 21, 22: 22, 23: 23, 24: 24, 25: 50, 27: 27, 38: 38, 53: 53, 82: 82}